Forums > Windsurfing   Gps and Speed talk

Garmin Fenix 7

Reply
Created by tbwonder > 9 months ago, 2 Feb 2022
tbwonder
NSW, 639 posts
2 Feb 2022 11:02PM
Thumbs Up

This thread is to discuss the new Garmin 7 Fenix. In particular if this watch could be considered for approval to GPSTC.
Apparently the sapphire ( and perhaps other) versions of this watch has multi-band technology. Which in theory provides far more accurate results.
Garmin have confirmed they are using Airoha as their chipset supplier (possibly similar to the AG3335M used in the Coros)

The Garmin manual states the following:
Auto Run -Enables the watch to detect ski or windsurf runs automatically using the built-in accelerometer. For the windsurf activity, you can set speed and distance thresholds for automatically starting a run.
SpeedPro -Enables advanced speed metrics for windsurf activity runs.

So lots of unanswered questions.
What are the advanced metrics that Garmin are talking about?
Is there some indication of the accuracy that can be logged?

BSN101
WA, 2244 posts
5 Feb 2022 5:29PM
Thumbs Up

$1000 watch tho. I want a watch to run the wsindsurding app so I can see categories while on the water but download with the motion.

Boston!
NSW, 248 posts
6 Feb 2022 7:57PM
Thumbs Up

Garmin charge what they can until people stop buying them, then adjust accordingly. They will be $500aud soon enough. Still, no error data, no GPSTC. If only they new about the 500 sales they are missing out on.

berowne
NSW, 1210 posts
8 Feb 2022 10:21AM
Thumbs Up

+1 I like having a smart watch for every day that also does GPS ( & GPSTC ideally).

Steve Charles
QLD, 1237 posts
8 Feb 2022 6:35PM
Thumbs Up

I have the Fenix 6 with downloaded GPS windsurfing software from Andy (tbwonder) Cannot recommend highly enough to get a Garmin to use along side your approved GPS device. There are a lot of people already doing this. I paid $600 for mine half price a year ago. Worth every cent.

Xbraun54
72 posts
10 Feb 2022 12:21AM
Thumbs Up

Select to expand quote
Boston! said..
Garmin charge what they can until people stop buying them, then adjust accordingly. They will be $500aud soon enough. Still, no error data, no GPSTC. If only they new about the 500 sales they are missing out on.


"Still, no error data" Based on what ??

segler
WA, 1597 posts
10 Feb 2022 10:37PM
Thumbs Up

I think it's based on gpstc's requirement for error data.

Since I don't have a need to post speeds to gpstc, I just use a cheap Timex Ironman GPS watch with a pretty good gps in it. Works fine and very simple to use. It displays speed, distance, and time of day on the water, and uploads to its own app and also to strava, which both show lots of results, including max speed and track.

aeroegnr
1477 posts
11 Feb 2022 12:29AM
Thumbs Up

My 6 (sapphire) has been bulletproof. Best watch I've owned honestly. I had a samsung smartwatch and it was buggy to the point of not being usable at times.

tbwonder
NSW, 639 posts
11 Feb 2022 9:59AM
Thumbs Up

I don't believe that the GPSTC has an actual requirement for "Error Data". In the past devices have been approved that do not produce it. The website does not mention that it is a requirement.
The current situation appears to be you can only post with one of the four approved devices all of which use doppler for speed measurement and record an error estimate. However as technology improves devices may come to market that are sufficiently accurate to be considered good enough for posting to GPSTC.
To date GPSTC have not specified how accurate a device needs to be, they simply assess each device on its merits.

So just for example IF this new Garmin watch can be tested along side a current approved device and consistently replicate the results, then I don't see why they would not approve it.

That is however a big "IF". Putting doppler and error data aside, There are many issues with wrist worn devices that have become apparent with the GW60. Which is why the software filters have had to be relaxed so much. But who knows perhaps this new multi-band technology will fix those signal reflection problems.

remery
WA, 1882 posts
11 Feb 2022 11:37AM
Thumbs Up

I wonder if tracks from 10 years ago come up with the same results when analysed with contemporary software.

GreenEgg
8 posts
11 Feb 2022 5:21PM
Thumbs Up

Not sure about the error data fields in the newest Garmin, but the Coros watches show all the error data needed. Just like the Motions. And in the end the doppler speeds are spot on between all the devices.


segler
WA, 1597 posts
11 Feb 2022 10:43PM
Thumbs Up

I believe Coros is also approved for gpstc. Where I get this is here:

www.gps-speedsurfing.com/default.aspx?mnu=item&item=GPSInfo and

www.gps-speedsurfing.com/default.aspx?mnu=item&item=gpsother&utm_source=seabreeze.com.au

ico
10 posts
11 Feb 2022 11:26PM
Thumbs Up

Select to expand quote


Gpstc is not the same as gps-speedsurfing.com.
gps speedsurfing approved the Coros devices.

Bigsnut
30 posts
12 Feb 2022 4:41AM
Thumbs Up

Based on that sample above the coros has deviation errors up to 2.7 units (kn or km it doesn't say).
I'm hoping the multi band chips are approved but I expect a lot of disappointed sailors haveing there screen data scores being filtered back after tracks are processed.

ico
10 posts
12 Feb 2022 3:21PM
Thumbs Up

Select to expand quote
Bigsnut said..
Based on that sample above the coros has deviation errors up to 2.7 units (kn or km it doesn't say).
I'm hoping the multi band chips are approved but I expect a lot of disappointed sailors haveing there screen data scores being filtered back after tracks are processed.





The image that GreenEgg showed is a piece of the data points from my tests. A small explanation: these are data points in Km/H. The Coros watches show the corrected Doppler speeds on the screen. These are the same numbers you see in GPS results. So no surprises afterwards.

The data in this table comes from an extreme test with a car. I firmly braked to trigger the error filter. You will understand that in such situations the filtering with a 1 hz device is more substantial than with measurements with a higher frequency.

What particularly caught my eye in this test was the difference between 5 hz and 10 hz in this table. The 5 hz Motion filters less than the 10 hz. This is probably explained by the fact that the 5 hz device had more satellites in sight at the time of testing than the Motion at 10 Hz.

What I find more important is the end result. I see that the 2 second peak with a 1 hz device like the Coros the deviations are logically greater at extreme (de)accelerations. The filter is then quite rigid and rounds - without exception - down. On 10 second runs, these differences almost completely disappear.
Over all 10 seconds of runs (almost 100 runs) , the maximum difference between the Motions and Coros Vertix 2 was once 0.53 km/h. The average difference was only 0.04 km/h (!!!).The chance that a device at 10 hz shows incorrect values is smaller, but in practice I come to the conclusion that the results are completely comparable.

ico
10 posts
12 Feb 2022 4:24PM
Thumbs Up

Table showing the differences between 10 sec runs Motion 10 Hz versus Coros Vertix2 1 Hz and GW60 5 Hz. Highlighted in white where the difference exceeds more than 0,25 km/h.






segler
WA, 1597 posts
12 Feb 2022 11:02PM
Thumbs Up

Thanks for the correction.

Aus501 Boz
WA, 110 posts
13 Feb 2022 9:47AM
Thumbs Up

Select to expand quote
ico said..

Bigsnut said..
Based on that sample above the coros has deviation errors up to 2.7 units (kn or km it doesn't say).
I'm hoping the multi band chips are approved but I expect a lot of disappointed sailors haveing there screen data scores being filtered back after tracks are processed.


The Coros watches show the corrected Doppler speeds on the screen. These are the same numbers you see in GPS results. So no surprises afterwards.



Interesting observations, I run the Coros Watch, Mini and GT31 at the same time during speed sailing sessions, the Coros is great for feedback on water but the data displayed can be out especially for the Alphas by 3-4 knots on the watch display, it does get corrected once posting to GPSSS but on water display is always out.

The 2 sec speed as you pointed out in high acceleration is not as accurate, the data does not seem to get corrected once posted. As an example yesterdays speed session Mini recorded 2 sec at 37.95, GT31 recorded at 38.05, Coros 37.09.

It is quite common the Coros reports minus 1-1.5 knots on the higher end of the spectrum. For GPSTC which also focuses on 2 sec speed this difference is quite large. As you have observed the difference as the time frame goes to 10sec is not as large and is comparable to the other devices.

I contacted Coros there reply below: Seems the data gets filtered at the Corros App level and I assume at the GPSSS level or one of the other software's that we use.

At least with the GW60 or GT31 you can with some confidence rely on the on water results, with the Coros great piece of hardware and great in all aspects of what it does but for now I think I'll only use it for on water feedback.

Jessie (COROS Wearables Help Center)Feb 11, 2022, 13:25 PST
Hi Nick,

We appreciate your patience. The data you see on the watch display is raw data. If there are changes in speed, the change in knots may be delayed for a few seconds. When syncing the watch to the app, there are algorithms and the workout data is parsed, filtered and recalculated. As a result, the information on the app provides the most accurate data compared to the watch.

Let me know if you have questions.

COROS Support Team

ico
10 posts
13 Feb 2022 4:51PM
Thumbs Up

@aus501 : about the Alpha indeed the numbers on screen are not the same as on GP3S. But the rest is as far as I know the same. Do you use the Vertix2?

Aus501 Boz
WA, 110 posts
13 Feb 2022 6:12PM
Thumbs Up

Select to expand quote
ico said..
@aus501 : about the Alpha indeed the numbers on screen are not the same as on GP3S. But the rest is as far as I know the same. Do you use the Vertix2?


Ico I use the ApexPro with the latest firmware, not disputing the numbers on the screen and what is reported to GPS3, the issue is when compared to the Mini/GT31 which correlate quite well even though one is 10Hz and the other 1Hz the Coros in nearly all occasions reports a lower 2 Sec sometimes this is up to 1-1.5 knots when directly compared to the Mini and GT.

Appreciate that GPS3 focuses on the 10sec but the GPSTC also focuses on the 2sec (Not that I can use the Coros for reporting to GPSTC, only my observations at present)

Not a tech guru so maybe it's all within acceptable parameters, just would be nice if what is on the water is actual especially for the Alphas.

ico
10 posts
13 Feb 2022 8:41PM
Thumbs Up

Yesterday we tested the Motion (1Hz), DIY Ublox device 10 Hz and 2x Vertix 1. I think all devices agreed pretty well (image showing 2S and 10 S).

But when the Going Gets Tough (like the Billy Ocean song), with big (de)accelarations or not perfect receiption of the satellites the Coros devices tend to round the numbers more down than the Ublox (and Locosys) devices. Fact is that the Ublox and Locosys Kalman filters react pretty much in the same way. We take it for granted that these are correct.



Aus501 Boz
WA, 110 posts
13 Feb 2022 9:49PM
Thumbs Up

Select to expand quote
ico said..
Yesterday we tested the Motion (1Hz), DIY Ublox device 10 Hz and 2x Vertix 1. I think all devices agreed pretty well (image showing 2S and 10 S).

But when the Going Gets Tough (like the Billy Ocean song), with big (de)accelarations or not perfect receiption of the satellites the Coros devices tend to round the numbers more down than the Ublox (and Locosys) devices. Fact is that the Ublox and Locosys Kalman filters react pretty much in the same way. We take it for granted that these are correct.




Thanks Ico, makes sense. The mini motion and GT31 I have mounted on the helmet so probably the best signal and less movement versus the watch over and underhand grip.The coros is an excellent piece of hardware heaps of battery life on GPS setting, trouble free and use it as a daily watch just not approved for GPSTC. Seems GPS3 has placed a lot of effort in getting another approved device, would be great if GPS3 could collaborate with the approvers of GPSTC to get a commercial device like the Coros watch approved for GPSTC competition as it seems a lot of the hard work verifying has been done by GPS3 already.

Aus501 Boz
WA, 110 posts
14 Feb 2022 10:15AM
Thumbs Up

Select to expand quote
ico said..
Yesterday we tested the Motion (1Hz), DIY Ublox device 10 Hz and 2x Vertix 1. I think all devices agreed pretty well (image showing 2S and 10 S).

But when the Going Gets Tough (like the Billy Ocean song), with big (de)accelarations or not perfect receiption of the satellites the Coros devices tend to round the numbers more down than the Ublox (and Locosys) devices. Fact is that the Ublox and Locosys Kalman filters react pretty much in the same way. We take it for granted that these are correct.





Ico another quick question, I use GPSResults V6.182 and when looking at the exported .fit file versus the .gpx file compared to the same session using the mini motion .oao file all analysed by the same software I find very close correlation to the results with the .gpx file to the .oao file as well the number of satellites show as 5 on the .fit file vs 11 on the .gpx file?

Which file is supposed to be analysed as the GP3S site states the .fit file?? Screen shot below .fit file, .Gpx file and the .oao file

EDIT: I think I've answered my own question the .gpx file doesn't come up with Doppler

Fit File

GPX File


.oao File



ico
10 posts
14 Feb 2022 4:00PM
Thumbs Up

Aus501 Boz: you answerd your own question correct .
Only 5 satelites... that is strange! I have never seen that before.... Normal between 10 and 14. Please test with other .fit files if this happens always.

I you want to see the error estimate of the datapoints in the .fit file: deselect the doppler box. You will get an extra field showing the difference between the data points and the final speeds.

I also made a manual for the Coros devices: www.surfcenter.nl/ingolf (link to the manual at the bottom of the text)




Aus501 Boz
WA, 110 posts
14 Feb 2022 6:46PM
Thumbs Up

Select to expand quote
ico said..
Aus501 Boz: you answerd your own question correct .
Only 5 satelites... that is strange! I have never seen that before.... Normal between 10 and 14. Please test with other .fit files if this happens always.

I you want to see the error estimate of the datapoints in the .fit file: deselect the doppler box. You will get an extra field showing the difference between the data points and the final speeds.

I also made a manual for the Coros devices: www.surfcenter.nl/ingolf (link to the manual at the bottom of the text)







Thanks ico, looked at 20 fit files all show 5 Sats.

As we sail at different sites up to 400klm apart I looked at each site and same thing 5 Sats. I find this strange as when I check in the Satellite systems showing number of Sats that are good and in view this number is always greater than 8.

Currently the watch mode is set on GPS,Glonass Galileo, QZSS. Completed a test using GPS,Beidou, Galileo,QZSS same result 5 Sats only.

Good manual my settings are all as per what you show in the manual so will be interesting to know why only 5 Sats are coming up in the analysis software.

Screen shot of watch sat signals




ico
10 posts
14 Feb 2022 7:03PM
Thumbs Up

I found the problem!!! It's in the newest version of GPS Results version V6.182. Version V6.181 is showing the correct number of sats..... Will report it to Manfred Fuchs.

Aus501 Boz
WA, 110 posts
14 Feb 2022 7:14PM
Thumbs Up

Thanks ICO, now I have to see what I have to do to get the GPSTC to at least look at the possibility of the Coros being approved for use

Tbwonder sorry for hijacking the thread with Coros but I think we are trying to do the same thing and have alternatives to what is on offer at present.

ico
10 posts
14 Feb 2022 7:16PM
Thumbs Up

Select to expand quote
tbwonder said..
This thread is to discuss the new Garmin 7 Fenix. In particular if this watch could be considered for approval to GPSTC.
Apparently the sapphire ( and perhaps other) versions of this watch has multi-band technology. Which in theory provides far more accurate results.
Garmin have confirmed they are using Airoha as their chipset supplier (possibly similar to the AG3335M used in the Coros)

The Garmin manual states the following:
Auto Run -Enables the watch to detect ski or windsurf runs automatically using the built-in accelerometer. For the windsurf activity, you can set speed and distance thresholds for automatically starting a run.
SpeedPro -Enables advanced speed metrics for windsurf activity runs.

So lots of unanswered questions.
What are the advanced metrics that Garmin are talking about?
Is there some indication of the accuracy that can be logged?




First test with the Fenix 7 in SpeedPro mode (=1 Hz)..... No doppler fields visible in GPS results and the results differ just as much as any other Suunto of Garmin. They need to put some more effort in the software to make it suitable voor speedsurfing.


ico
10 posts
23 Feb 2022 2:58AM
Thumbs Up

New version of GPSResults V6.185 with the fix of number of sats for the Coros .fit files.

Sepp
24 posts
25 Feb 2022 3:05PM
Thumbs Up

Tests in a german windsurf forum showed that the use of Glonass (in combination with all the other possible systems) the Coros will measure signifiantly lower speeds. Up to 5 kmh in GPS results.

GPS-Speedsurfing:...In general it's good to use the GPS/QZSS setting (can be changed SYSTEM, MORE SETTINGS, MODE...)

Aus501 Boz
WA, 110 posts
25 Feb 2022 4:41PM
Thumbs Up

Select to expand quote
Sepp said..
Tests in a german windsurf forum showed that the use of Glonass (in combination with all the other possible systems) the Coros will measure signifiantly lower speeds. Up to 5 kmh in GPS results.

GPS-Speedsurfing:...In general it's good to use the GPS/QZSS setting (can be changed SYSTEM, MORE SETTINGS, MODE...)


Thanks Sepp But in the Apex Pro only three possible settings below no ability for GPS/QZSS only

GPS
GPS,GLINASS,GALILEO,QZSS
GPS,BEIDOU,GALILEO,QZSS



Subscribe
Reply

Forums > Windsurfing   Gps and Speed talk


"Garmin Fenix 7" started by tbwonder